Match your song with the top 1,000 playlists on Spotify

Do you have a popular playlist on Spotify with many requests for adding new tracks? Are you a music label spending hours listening to new demos? How do you filter the demos you receive?

Matching artists with music labels has always been one of the major pains in the music industry: Several musicians are looking for a way to get promoted, get famous and finally earn money with their music, while only a few labels are willing to promote them.

Artists and labels are like job applicants and companies. One needs the other, yet without a clear methodology, the risk to waste time and energy looking for each other is enormous.

The same happens in music: Music creators struggle to find the right label that is willing to promote their new song.

Music labels, on the other hand, receive tons of demos per day and are left with the dilemma of spending time (hence money) listening to all the demos, or just rely on their network. While the latter seems the best solution of established labels, the first way is yet a good approach for new emerging publishers.

Any alternative?

In the last years, various alternatives evolved to emancipate artists from labels and publishers, here I list the 2 most popular ones:

  • Crowdfunding
  • Spotify

Crowdfunding, in the form of Kickstarter or other platforms, allows the artist to get funded for a new album, but it does not solve, in my opinion, the delicate dependency between artist and labels (or publishers).

Spotify

Spotify, on the other hand, is upgrading the relationship artist-label to a new one: artist-playlist. With Spotify, it is not the label in charge for promoting the song. A Spotify track is itself the main source of income for the artist. Anyone can upload a track on Spotify and earn money based on listening. The intermediary step with labels or publishers are not adding any value to the chain, hence those figures are no longer required (up to an intermediate level).

Uploading a song to Spotify does not create money though, unless the number of listens is high enough. Spotify pays $0.006 to $0.0084 per stream to the holder of music rights. This results in a race for listens: the more streams the more money earned.

How to generate a high number of listens?

One of the most powerful ways to make money on Spotify is to add a track on a popular playlist.

How many playlist’s owners are willing to include your song in their playlist?

Playlist owners, like labels, have tons of requests of people asking them to include tracks on their playlist. They are left to the same label’s dilemma of listening to all requests or relying on their private network. The same labels dilemma has just been translated to Spotify. How can playlist owners filter the most compatible tracks for his playlist?

Are you the only artist asking for this request?

Of course not. That is why playlists charge for this kind of service as well, like labels charge for promotion. How to convince a popular playlist’s owner to include your track in his playlist?

Imagine if matching tracks with playlists was done instantaneously, without listening to the tracks.

At Clockbeats, we developed an algorithm that matches tracks with playlists, without spending time.

How does it work?

Given a specific playlist, we are able to analyse all the songs belonging to that playlist, according to various parameters, e.g. Danceability, Energy, Tempo etc. An accurate analysis of all the playlist’s tracks is then used to identify important characteristics of the playlist itself, for various parameters. An example is given in the Figure below, representing the distribution of 6 parameters for all the tracks in a playlist.

Analyis of a playlist — The plot shows the distribution of the value of 6 parameters of each track of the playlist

As outcome of this analysis, several insights are retrieved, for that playlist.

Is this playlist homogeneous with regards to the types of track it contains? Is it in line with the trends for a specific music genre? Are the tracks of the playlist in the best possible order?

Those are examples of questions that we are able to answer with great accuracy thanks to a proper analysis of its music properties, without having to listen to all the songs nor depending on human bias.

How to match a track with a playlist?

A similar analysis as above can be done for a single track. Once both Playlist and Track are analyzed, it is possible to compare them and evaluate a possible fit. In the Figure below, a simple example is given to understand the matching algorithm. The red star represents the value of the parameters for the track; The box plots identify the playlist, for each audio parameter. A compatibility match can be assessed looking at the relative position between the red star and the box plot. If the red star lays far away from the box, the match is weak.

Matching a playlist with a new track — The red star represents the values of the new track, the box shows some statistical indicators of the playlist. The position of the red star against the playlist box gives some indications about the match between playlist and track.

A further analysis of the results allows us to understand the level and type of compatibility between playlist and track.

As final outcome of the analysis:

  • The playlist’s owner can quickly identify tracks with low level of compatibility and dive deeper in the analysis of tracks with higher level of compatibility.
  • The artist can use this tool to avoid wasting time contacting the less compatible playlists. Moreover, given a playlist, an artist can create a new good matching track that a playlist owner cannot refuse.

Try out our demo at https://spotimatch.com/.

The core algorithm can be found here: https://github.com/nicolomantini/SpotiMatch